The influence of the peat composition on the whisky flavours

Introduction:
In 2007, I wrote a first article about Whisky, Peat and Phenols. More recently, in 2009, BM Harrison and FG Priest have published an interesting article on the composition of peat in the preparation of Scotch Whisky Production and the influence of the geographical source and extraction depth. (Composition of Peats Used in the Preparation of Malt for Scotch Whisky Productions-Influence of Geographical Source and Extraction Depth, Barry M. Harrison and Fergus G. Priest, J. Agric. Food Chem. 2009, vol 57, N° 6, pp. 2385-2391).

Material and methods:
Peat samples from four locations (Islay, Orkney, St. Fergus (Aberdenshire), and Tomintoul (Speyside) were analyzed using Curie point pyrolysis in combination with gas chromatography-mass spectrometry (Py-GC-MS). In addition, samples from Orkney and Islay (Glenmachrie) were taken at different depth. At Orkney, five samples were taken; from just below the surface to the bottom of the bank at 30 cm intervals and at Islay, four samples were taken, from about 10 cm below the surface to about 30 cm from the bottom of the bank, at 25 cm interval for evaluating the effect of depth on peat composition. Briefly, the method consisted in burning (i.e., pyrolysis) the peat at 610°C and the burnt product (i.e., pyrolysate) was analysed by a method allowing the separation and identification of the different constituents (i.e., GC-MS). For the full details, refer to Composition of Peats Used in the Preparation of Malt for Scotch Whisky Productions-Influence of Geographical Source and Extraction Depth, Barry M. Harrison and Fergus G. Priest, J. Agric. Food Chem. 2009, vol 57, N° 6, pp. 2385-2391).

Results:
In total, out of the 106 products identified, and 92 compounds were having a significant effect on the separation of the four geographical locations. The compounds were broadly split into the following classes: phenolic compounds, carbohydrate derivatives (“sugars”), aromatic compounds, and nitrogen-containing compounds. Using statistical methods (principal component analysis (PCA)), the ratio of phenol derivatives (carbohydrate derivatives to guaiacols, syringols, and phenols) was the major discriminator between the samples of the different geographical regions, explaining more than 60% of the variance. St. Fergus and Islay samples were characterized by high percentages of guaiacols (aroma described as aromatic, phenolic, burnt, woody, bacon, savoury, smoky, and medicinal), syringols  (aroma described as aromatic, phenolic, spicy, smoky, baconlike, sweet, medicinal, creamy, meaty, and vanilla), and phenols (aroma described as aromatic, phenolic, burnt, woody, bacon, savoury, smoky, and medicinal) in the pyrolysate. Relatively high proportions of carbohydrate derivatives in the pyrolysate characterized Tomintoul and Orkney samples. Furthermore, the samples from Tomintoul were found to yield a significantly higher proportion of carbohydrate derivatives than the samples collected from Orkney.

Impact of Extraction Depth on Composition.
A decrease in the carbohydrate content with increasing depth was detected, with the greatest reduction in carbohydrate content occurred in the upper 60 cm of the Orkney depth profile. Below this, carbohydrate losses were relatively small. With the reduction in the proportion of carbohydrates at increasing depth, there was a general increase in the proportions of the other classes of compound. In the case of the Islay depth profile, the carbohydrate loss was greatest in the upper 25 cm. The proportion of guaiacols, phenols, and syringols increased with increasing depth.
The PCA analysis of these results showed that, in the case of the Orkney depth profile, extraction depth had a relatively large impact on composition. The younger peat samples taken from the top of the depth profile, were similar to the Tomintoul samples. Older peat samples were similar to Islay samples. The Islay depth profile samples were less varied in composition than the Orkney samples.
For the full details, please refer to the original publication.

Conclusion:
The Authors concluded that the origin of the peat and depth resulted in different chemical profiles once burnt.  The impact of these observations is limited, since this reflects the changes in the “aromatic profile of the smoke”, but how is it relevant to the final product, the whisky? 
To answer that question, we have to go through Barry Harrison Thesis: Peat source and its impact on the flavour of Scotch whisky, Heriot-Watt University, 2007.

Thesis
Once it was demonstrated that the source of the peat had an impact of the chemical composition of the peat smoke, B. Harrison investigated if these differences would impact the flavours of the whisky distillate.

Material and Method:
Since it is not possible to conduct such experiments in a working distillery, the first task of B. Harrison was to create a laboratory-scale peating process in order prepare malted barley kilned with peat from the different sources mentioned above. This malt was then mashed, fermented and distilled in order to recreate the industrial whisky making process. Finally, the distillate was analyzed using analytical equipment (Gas chromatography coupled to mass spectrometry (GC-MS), as well as by sensory analysis (by nosing). Of note, distillation was performed using glassware and not copperware.
Finally, commercial peated malt and new make spirits were compared with the results from the laboratory-scale process.

Results:
Malting/kilning:
The lab-scale peating process resulted in malt peated to higher levels than industrial peat malts, as well as different proportions of phenols, in particular the famous guaiacols. Furthermore, chemical analysis of the peated malt indicated that classes of compounds identified in the pyrolysate of the peat (see first section of this article) could also be detected in the peated malt.

Comment:
The process of a laboratory-scale peating might appear as a straightforward process, but the reality was very different (think about the pagoda roofs of Charles Doig) and B. Harrison did a remarkable work in trying not only to recreate a working “kiln”, but also to develop the whole process as well as robust analytical setup in order to simulate as close as possible the commercial process and product and to identify the transfer of the peat component from the smoke to the final product. Importantly, the different types of aromatic compounds present in the burnt peat could also be identified in the malted barley (meaning that compounds present in the peat are transferred to the malted barley).

New Make Spirit:
Different types of phenols could be identified in the lab-scale produced new make spirit, elaborated with malt peated with peat from different origins. The lignin derivate compounds (“wood” or compound of the cell wall in plants) present in the peat were not transferred to the spirit. A total of 26 phenolic compounds were identified by chemical characterisation of the lab-scale peated new make spirit.

Comment:
New make using different peat resulted in new makes with different chemical profiles (e.g., amount of sugar, amount and proportion of different phenolic compounds).

Sensory analysis of new make spirit:
Trained analyst (panellists) tried to identify the characteristics of the different new make spirits. The so-called “attribute” (type of flavours) sweet, cereal or fruity/estery were not different between them. However, the green/grassy attribute was different (statistically significant) between those samples, with the highest values for the samples from Orkney and the lowest ones from the different Islay samples.
Comparison between lab-scale and industrial new make samples indicated that the burnt, smoky and medicinal peaty flavours were different, with the industrial samples being more medicinal while the lab-scale samples were smokier and more intense burnt flavours.

Comment:
New makes prepared using peat of different origin contained compounds that could separate the new-makes according the origin of the peat source. However, different chemical composition does not always translate in different flavours or aromas that can be perceived by the final consumer. By performing sensory analysis, trained analysts smelled the different samples to check if our nose would be able to detect any relevant olfactory differences.

Discussion:
Peat is commonly used in Scotland for the production of single malts, not only on Islay, but also on Orkney, in the Highlands, Campbeltown as well as in the Speyside, with peat extracted closed to the distillers or malters. The climate in these regions is different as well as the vegetal composition, thus resulting in peat derived from different organic origins (different vegetal). In the whisky production, peat is used to kiln (“peat”) the barley and the malted barley will then be mashed, fermented and double distilled. For kilning, the bottom layers in the peat bogs are used and the results of this study indicated a significant difference between the top and the bottom layers. Since the middle and bottom layers are similar, therefore cutting depth should have only little impact on the flavour profile of the malted barley.
Kilning is an important step and the flavours present in the peat smoke are transferred (partially) to the barley. The work of Harrison indicated that the burning temperature of the peat had a significant impact on the chemical composition of the peat smoke. For experimental purposes, one unique temperature was used for all samples. Not only temperature but also moisture play a significant role. Therefore, a distiller could play with these parameters to change the aromatic profile of his whisky. In the distillate (new make), the origin of the peat could be clearly identified by analytical methods. By sensory analysis (by “human nose”), the spirits using Tomintoul (Speyside) peat was more medicinal than the sprit using Hobbister (Orkney) peat. The level of peat aroma was low in the St-Fergus spirit (Aberdeenshire) despite a high abundance of aromatic peaty aromas, but the spirit was sweet, spicy and medicinal.
Differences in the malted barley and new make spirits between lab-scale and industrial-scale production were observed. In spite of this disparity, one can safety conclude that the origin of the peat has a significant impact of the flavours of the new make spirit and ultimately on the final product.
The authors also compared the peat extracted from 3 different locations on Islay. From a statistical point of view, the differences between them were not as pronounced as between the “Islay group” and the others. I would be interested in comparing these differences from a sensory point of view. During one of my discussions with John Campbell, distillery manager at Laphroaig a few years back, a batch of barley malted 100% on their own floor maltings using exclusively peat cut in their own peat bog was distilled in the early 2000s. I was told that the new make was different and could not get more information. This might be the “ultimate” Laphroaig to be tasted during the Open Day at Laphroaig during the Feis Ile 2011.

Conclusion:
The work of Harrison indicates that the origin (source) of the peat has an impact on the flavour of the spirit.
Slainte and thanks to BM Harrison for providing me with a copy of his publication.

Dr P. Brossard, 29 May 2011